Understanding students' evaluations of professors using non-negative matrix factorization

In this paper, we use Nonnegative Matrix Factorization (NMF) and several other state of the art statistical machine learning techniques to provide an in-depth study of university professor evaluations by their students. We specifically use the Kullback-Leibler divergence as our loss function in keeping with the type of the data and extract revealing patterns consistent with the educational objectives underlying the questionnaire design. In particular, the application of our techniques to a dataset gathered at Gazi University in Turkey reveals compelling patterns such as the strong association between the student's seriousness and dedication (measured by attendance) and the kind of scores they tend to assign to the courses and the corresponding professors. Insights emerging from our study suggest that more aspects of students' evaluations should be explored at greater depths.